Commit Graph

3971 Commits (29d7a2a5589bf8120ca6a17d5d74aaa0e76d1644)

Author SHA1 Message Date
Yicheng Qin 70b501d17c Merge pull request #1824 from yichengq/241
etcdserver: close idle connections when stop sendhub
2014-12-02 10:12:45 -08:00
Xiang Li 6692a8060e Merge pull request #1823 from xiang90/raft_log
raft: logging state change events and events on bad path
2014-12-02 10:10:08 -08:00
Xiang Li 51de095d2c raft: logging state change events and events on bad path 2014-12-02 10:08:19 -08:00
Xiang Li f02eae934b Merge pull request #1827 from lamielle/doc-whitespaces
doc: uses spaces consistently in sample JSON
2014-12-02 07:44:03 -08:00
Alan LaMielle 2b7af3d101 doc: uses spaces consistently in sample JSON
Replaces spurious uses of tabs with four spaces.  Removes strange
unicode space character with standard space character.
2014-12-02 06:43:07 -08:00
Yicheng Qin aa61009560 wal: not return ErrIndexNotFound in ReadAll
This IndexNotFound case is reasonable now because we don't write dummy
entries into wals any more.
2014-12-02 00:28:54 -08:00
Yicheng Qin fa292391d8 etcdserver: close idle connections when stop sendhub 2014-12-02 00:08:47 -08:00
Yicheng Qin f34fe6e4ae Merge pull request #1819 from yichengq/239
integration: use timeout transport when launching cluster
2014-12-01 23:00:28 -08:00
Xiang Li cb74b6812b Merge pull request #1820 from xiang90/fix_storage
raft: fix memory storage
2014-12-01 21:25:17 -08:00
Xiang Li 312db7f0f3 raft: fix memory storage
Memory storage should append all entries that have greater index
than the snap.Matedata.Index. We first truncate the old parts of
incoming entries. Then truncate the existing entries in the storage.
At last, we append the incoming entries to the existing entries.
2014-12-01 16:37:16 -08:00
Yicheng Qin 7a1d147795 integration: use timeout transport when launching cluster
This makes it do the same behavior as etcdmain does.
2014-12-01 16:26:27 -08:00
Xiang Li 19ccdbee18 Merge pull request #1806 from xiang90/no_copy
No copy
2014-12-01 13:15:13 -08:00
Xiang Li 7beac083ff Merge pull request #1810 from xiang90/purge
*: support purging old wal/snap files
2014-12-01 12:05:05 -08:00
Xiang Li d3db010190 *: support purging old wal/snap files 2014-12-01 11:50:17 -08:00
Xiang Li 92d4112feb Merge pull request #1809 from xiang90/unstable
raft: stableTo checks term matching
2014-12-01 11:09:40 -08:00
Xiang Li 649176934a raft: add tests for stableTo 2014-12-01 10:54:34 -08:00
Xiang Li 7c47decd19 Merge pull request #1813 from xiang90/snap_event_log
etcdserver: log snapshot event
2014-11-30 12:11:23 -08:00
Xiang Li bc5acd3c42 etcdserver: log snapshot event 2014-11-30 12:10:20 -08:00
Xiang Li 3c0fbe285c raft: stableTo checks term matching
stableTo should only mark the index stable if the term is matched. After raft sends out unstable
entries to application, raft makes progress without waiting for reply. When the appliaction
calls the stableTo to notify the entries up to "index" are stable, raft might have truncated
some entries before "index" due to leader lost. raft must verify the (index,term) of stableTo,
before marking the entries as stable.
2014-11-28 14:13:07 -08:00
Xiang Li d214e87aee raft: make unstable.entries immutable; copy the entries at bad path 2014-11-27 19:35:03 -08:00
Xiang Li d244e3bf6e raft: fix node bench 2014-11-26 23:07:35 -08:00
Xiang Li fe0bc4ff36 Merge pull request #1805 from xiang90/fix_raft_b
raft: fix start term
2014-11-26 21:41:38 -08:00
Xiang Li 746c66b466 raft: fix start term 2014-11-26 21:21:13 -08:00
Xiang Li 35cf7b5a31 Merge pull request #1800 from xiang90/unstable
raft: move unstable related function to log_unstable.go
2014-11-26 16:12:43 -08:00
Xiang Li 7929e46dd8 raft: clean up 2014-11-26 15:31:07 -08:00
Xiang Li 8a626257c7 raft: move unstable related function to log_unstable.go 2014-11-26 15:25:24 -08:00
Yicheng Qin 416b799ecf Merge pull request #1788 from yichengq/233
rafthttp: increase the size of streaming buffer
2014-11-26 15:22:28 -08:00
Yicheng Qin 00ce0702b9 rafthttp: increase the size of streaming buffer
Streaming buffer is used for:
1. hand over data to io goroutine in non-blocking way
2. hold pressure for temprorary network delay
3. be able to wait on I/O instead of data coming under high throughput

The old 1024 value is too small and is very likely to be full and
break the streaming when suffering temprorary network delay.
2014-11-26 14:46:52 -08:00
Yicheng Qin 7358ef21a2 Merge pull request #1799 from yichengq/237
integration: attempt more times to listen on specified port
2014-11-26 14:26:41 -08:00
Yicheng Qin e03cf6d488 Merge pull request #1797 from yichengq/236
raft: no need to save dummy entry into stable storage
2014-11-26 14:23:32 -08:00
Yicheng Qin 670d98ec72 integration: attempt more times to listen on specified port
Travis is rather slow, and it may fail to listen on that port sometimes.
2014-11-26 14:21:15 -08:00
Yicheng Qin 0f070f3d2d raft: no need to save dummy entry into stable storage 2014-11-26 14:04:56 -08:00
Xiang Li b2d686495c Merge pull request #1796 from xiang90/unstable
raft: move all unstable stuff into one struct for future cleanup
2014-11-26 13:59:07 -08:00
Xiang Li 66252c7d62 raft: move all unstable stuff into one struct for future cleanup 2014-11-26 13:36:17 -08:00
Xiang Li 488f508505 Merge pull request #1777 from xiang90/log_interface
Log interface
2014-11-26 12:51:37 -08:00
Yicheng Qin ab2a40ea37 Merge branch 'log_interface'
Conflicts:
	raft/log.go
2014-11-26 12:16:02 -08:00
Xiang Li 732cfa1ad6 raft: remove the applysnap from Storage interface 2014-11-26 11:28:51 -08:00
Xiang Li e23f9e76d1 raft: do not applysnapshot in raft 2014-11-26 10:59:13 -08:00
Xiang Li d01d6119e5 Merge pull request #1792 from xiang90/fix_raft_l
raft: always write dummy entry to storage
2014-11-25 23:32:01 -08:00
Xiang Li 39e6631447 raft: always write dummy entry to storage 2014-11-25 23:27:40 -08:00
Yicheng Qin 7614aa53bf version: bump to alpha.4 2014-11-25 21:41:13 -08:00
Yicheng Qin 006da2f8a0 Merge pull request #1789 from yichengq/234
rafthttp: only batch good MsgAppResp
2014-11-25 18:46:23 -08:00
Yicheng Qin d5ceb26408 rafthttp: only batch good MsgAppResp
A MsgAppResp with Reject set should be sent back to the leader as soon
as possible instead of batching.
2014-11-25 17:53:26 -08:00
Xiang Li 8de98d4903 raft: clean up 2014-11-25 16:21:50 -08:00
Xiang Li 9bd1786fe4 raft: memory storage does not append out of date entries 2014-11-25 15:18:40 -08:00
Xiang Li 9df0e7715d raft: do not panic on out of date compaction 2014-11-25 15:14:39 -08:00
Xiang Li 01cbcce8ba etcdserver: do not applySnapshot twice 2014-11-25 14:53:49 -08:00
Xiang Li 74d8c7f457 etcdserver: cleanup main loop 2014-11-25 14:38:18 -08:00
Yicheng Qin 7e6e305c4f Merge branch 'log_interface'
Conflicts:
	raft/raft.go
2014-11-25 14:22:11 -08:00
Yicheng Qin a13d5a70ff etcdserver: save snapshot before entries 2014-11-25 12:39:15 -08:00